training AI for hardware design

We’re hardware engineers who got tired of waiting for better AI tools.

We’re building reinforcement learning environments that teach models to write RTL, build testbenches, and debug designs.

Hardware design is full of open-ended problems that models should be able to tackle. But training data is scarce, and its quality is uneven. Our environments give models practical engineering tasks to learn from, with rewards based on whether their designs meet the spec.